The stability of discrete Hopfield neural networks not only has an important theoretical significance, but also is the foundation of the applications of the networks. Security of the algorithm is discussed.ANN originated has been applied in all aspects of life and production, especially in intelligent control, nonlinear function approximation and environmental parameters forecast assessment. Be similar to the human's brain mechanism, a neural network complete the tasks by the process of learning (training) and use (associative memories). Normally, the neural network can be assured used only after thorough study. According to the dynamic characteristics of the interconnections between nerve cells in Hopfield network, this paper discusses the application of Hopfield network model in gene diagnosis, points out the network problems and puts forward the solutions.